[gmx-users] Membrane Proteins & pdb2gmx

[Peter C. Lai]

On 2011-12-08 02:43:13AM -0600, Anirban Ghosh wrote:
> Hello Justin,
> 
> In your membrane protein simulation tutorial after making the topology, you have mentioned that "Placing the new gromos53a6_lipid.ff directory in $GMXLIB will allow you to use this force field system-wide." I suppose this is valid only for the proteins (and not membranes) to be processed through pdb2gmx using gromos53a6_lipid force-field, right? And to process a membrane using pdb2gmx we need to change the aminoacids.rtp file with the relevent POPC/DSPC/DPPC etc. entries. Right? Or can we somehow make pdb2gmx use the POPC/DPPC/DSPC.itp file?

Yes because in that tutorial, you are not adding an rtp with the lipid
residue names to generate lipid topologies "from scratch". The tutorial
uses pregenerated lipid topologies from Tieleman. You are certainly
able to add your own rtp files; pdb2gmx searches all validly formatted
rtp files in $GMXLIB/forcefield.ff/ for topology information by residue 
name. For example, when I converted parts of CGenFF to gromacs, I added
my own library of residue types for the small molecules I am modeling as
its own .rtp file. (Basically what I mean is, there is no need to hack
aminoacids.rtp to add new residues, just add your own .rtp file).

However you obtain a topology of a moleculetype (either using pdb2gmx or 
from someone else e.g. Tieleman) with a given forcefield, you do not need to 
run pdb2gmx again to generate another topology using that forcefield for 
subsequent tasks. Just include the relevant .itp file in the .top file 
you specify to grompp. For example, I generated POPC.itp once and have been 
using the same file for all my simulations involving POPC. 

Here is an entire membrane protein+lipid+water+ligand .top ready for
grompp -p (I use charmm36 with cgenff added - this is analogous to Justin's
gromos53a6_lipid modification):

----
#include "charmm36cgen.ff/forcefield.itp"
#include "gpcr.itp"
#include "popc.itp"
#include "charmm36cgen.ff/tips3p.itp"
#include "charmm36cgen.ff/ions.itp"
#include "ligand.itp"

[ system ]
; Name
GPCR in POPC with LIGAND

[ molecules ]
GPCR             1 
; the gpcr.itp file starts with:
;   [ moleculetype ]
;   ; Name            nrexcl
;   GPCR             3
POPC              230
SOL             21468
NA                  2
LIGAND              1
----

gpcr.itp, popc.itp, and ligand.itp were all originally generated by me using 
pdb2gmx on the relevant molecules coordinate files at one point or another.
Again, these individual .itp files are reusable for other simulations 
involving the same molecule types, protonation states, and charmm36cgen.ff.

Hope that helps.

-- 
==================================================================
Peter C. Lai			| University of Alabama-Birmingham
Programmer/Analyst		| KAUL 752A
Genetics, Div. of Research	| 705 South 20th Street
pcl at uab.edu			| Birmingham AL 35294-4461
(205) 690-0808			|
==================================================================